Juvenile IGF-I: an update

Transcription

1 Juvenile IGF-I: an update Kim Bunter and Uwe Wuensch Animal Genetics and Breeding Unit (AGBU), a joint venture of NSW Agriculture and the University of New England, University of New England, Armidale, NSW, 2351; Pig Breeding Association of Middle Germany, PSF 954, Chemnitz, 09009, Germany. Introduction In past Pig Genetics Workshops, we have provided research results on the relationships between juvenile IGF-I and piglet or performance traits, and discussed how to use this information in your breeding programs. The purpose of this document is simply to update pig breeders on recent developments, including new parameters from a trial conducted in Germany by Mitteldeutscher Schweinezuchtverband e. V on German Landrace. Developments 1. Default genetic parameters for IGF-I have been updated in PIGBLUP. 2. Pigmania, AGBU, and Primegro Limited have cooperated to develop an automated data entry system for IGF-I test results into the herd management software, and generate PIGBLUP data files with appropriate contemporary group coding for this trait. MIPS also have facilities to incorporate IGF-I test results into PIGBLUP data files. 3. Direct use of assay codes (provided by Primegro Limited in result files) in PIGBLUP for contemporary group formation is currently under development. 4. Herdsman (S & S programming, Indiana, United States) are incorporating IGF-I data into their integrated software for record keeping and genetic evaluation. 5. Licensee s of the technology have significantly increased. The total number of tests conducted for juvenile IGF-I in the past 12 months was approximately Analyses for new trial data obtained from Mitteldeutscher Schweinezuchtverband e. V. (German Landrace) has been completed. 7. Research into the use of IGF-I testing for cattle breeding is ongoing. AGBU Pig Genetics Workshop March

2 Trial results for Mitteldeutscher Schweinezuchtverband e. V. (MSZV) 1. Description of Trial IGF-I data Blood spot samples from 650 German Landrace (GL) animals were collected by Mitteldeutscher Schweinezuchtverband e. V. during the period of 5/12/2000 to 27/12/2001, inclusive, under a PrimeGRO IGF-1 trial license. Samples were collected on 48 separate occasions (test dates) with between 1-30 animals tested per occasion (average: 13.5 animals/test date). Nearly all GL animals entering the MSZV station for performance testing were bled. However, there were no individuals whose sire had also been tested for IGF-I as a juvenile. At the completion of all performance testing (i.e. all sample cards were stored) Primegro assayed cards on three occasions, giving rise to nine assay groups in total (average: 72.2 animals/assay group, which incorporates assay date and assay plate). All blood spot samples collected were assayed regardless of whether pigs completed performance testing or not. Performance data Performance data were recorded for a large number of traits by MSZV. All German Landrace males that entered the test station during the IGF-I trial period were included in the performance data file, regardless of whether they commenced (at 30 kg) or completed (at 110 kg) performance testing. Thus, performance data was available for 786 animals; progeny of 49 sires and 306 dams, from 369 litters. Because of the extended trial period, several dams had progeny performance tested from more than one litter. Moreover, the use of AI sires provided links across the two farms from which progeny were sourced. The total number of animals in the pedigree was Analyses focussed on traits for which EBVs are routinely produced (marked with *), along with some other traits that may be of interest. Traits for which parameter estimates were obtained include: *FI_TOT: total feed intake during performance test (kg) ADG2: daily gain during performance test (gm/day) *ADG3: lifetime average daily gain (gm/day) FCR: feed conversion ratio (kg feed/kg gain) *US_BF: ultrasound backfat thickness (mean of three measures: mm) *M_AREA: carcase muscle area (cm 2 ) *BF_AREA: carcase fat area (mm 2 ) *PH1_K: ph at 1 hour post slaughter IMF: intramuscular fat (%) BELLY: belly score (1-9 scale) Traits used as linear covariates in models for analyses were: US_WT: weight when ultrasonic back fat measures were taken (kg) END_WT: weight at the end of performance test (kg) HCC_WT: hot carcase weight (kg) AGBU Pig Genetics Workshop March

3 The distribution of data for each trait (including covariates) was examined with PROC UNIVARIATE (SAS Institute) to identify outliers. For the majority of performance traits, no outliers were evident. However, values for END_WT were treated as missing if <98 kg or >127.8 kg; values for FCR were treated as missing if >3.4; and values for IMF > 2.0 were also deleted as outliers. Performance records for animals with missing design variables (eg. missing date of birth, slaughter date etc), or with covariates identified as outliers, were not included in analyses. Models used for analyses Examination of the IGF-I data showed that animals were not sampled for serum IGF-I levels at a consistent age throughout the entire trial. There were three distinct groups, with peak frequencies at days, days, and 48 days. Group codes were thus generated to distinguish, at least approximately, between animals measured at different ages and/or with different intervals from weaning to sampling. Groups were coded for the intervals days, days, and days, based on the frequency distribution for age at sampling. Animals outside the age range of days were excluded from the data. Outliers with IGF-I levels exceeding 450 ng/ml were also removed, leaving 576 animals after editing with records for IGF-I. Estimates of genetic parameters for all animals present in the pedigree were obtained using ASREML for the edited IGF-I data. Fixed effects included test year and month, assay group, and group for age at testing. Age, fit as a linear covariate, was not significant when group was included in the model. Random terms included animal and litter effects, giving rise to estimates of heritability and common litter effects. MSZV provided the models used in the estimation of breeding values for FI_TOT, M_AREA, BF_AREA, PH1_K, ADG3 and US_BF. However, these models are used for analyses that combine data from several test stations and abattoirs, along with data from on-farm performance recording. Thus, some of the model design variables are redundant for the analysis of this data (eg. test station, sex, abattoir). Systematic effects accounted for when estimating genetic parameters for each performance trait are shown in Table 1. Table 1 Systematic effects for performance traits Trait Farm PJM FJQ SJM SOZ US_WT END_WT HCC_WT FI_TOT ADG2 FCR ADG3 US_BF M_AREA BF_AREA PH1_K IMF BELLY farm: farm of origin (2 levels) PJM: is year month of test, based on performance end date (17 levels) FJQ: is farm year quarter, based on performance end date (13 levels) SJM: year month of slaughter, based on slaughter date (17 levels) SOZ: is slaughter date (66 levels) AGBU Pig Genetics Workshop March

4 For each trait, parameter estimates were obtained under an animal model (univariate analyses). Litter was fitted as additional random effect for ADG2, ADG3 and US_BF, and was also significant for PH1_K. However, MSZV do not include litter effects when estimating breeding values for PH1_K. Estimates of the correlations between IGF-I and the above performance traits were obtained from a series of bivariate analyses. 2. Results and Discussion Raw data characteristics Characteristics of the raw performance data, prior to editing, are presented in Table 2, along with characteristics of the IGF-1 data. Coefficients of variation (CV) tended to be relatively low for most traits, particularly PH_1K. The exceptions were US_BF and BF_AREA, with moderate coefficients of variation, and IMF and BELLY, which were highly variable relative to the mean value. The CV for juvenile IGF-1 was also large, reflecting the multiple age groups noted earlier. Table 2 Characteristics of the raw data prior to editing (SD: standard deviation; Min: minimum value; Max: maximum value; CV: coefficient of variation) Trait N Mean SD Min Max CV IGF-I FI_TOT ADG ADG FCR US_BF M_AREA BF_AREA PH1_K IMF BELLY Estimates of genetic parameters Parameter estimates from univariate analyses are presented in Table 3, along with the number of records used and trait means after editing (described above). Editing procedures had the greatest impact on mean values for ADG2, which was 8 g/day higher in the edited data set. However, this change is not significant. Estimates of heritabilities ranged from 0.29±0.15 (IGF-I) to 0.89±0.12 (IMF). The magnitudes of heritability estimates are generally consistent with those reported for the same or similar traits elsewhere. MSZV use heritabilities of 0.45, 0.37, 0.66, 0.56 and 0.09 for FI_TOT, US_BF, M_AREA, BF_AREA and PH1_K when estimating breeding values. Although the heritability estimated from trial data is lower for FI_TOT, and higher for PH1_K, heritability estimates for the remaining traits are in very good agreement with the reported values. Estimates for IMF are very high, and the models AGBU Pig Genetics Workshop March

5 used for analysing this trait may require further development. For all traits standard errors are large due to the relatively small amount of data used. Estimates of correlations between juvenile IGF-I and the performance traits are shown in Table 4. Genetic correlations between juvenile IGF-I and FI_TOT, FCR and US_BF were very high, positive, and significantly different from zero. This outcome is consistent with results from previous trials. Similarly, the genetic correlation between IGF-I and BF_AREA was high and positive. This would be expected given the very high (cv 0.74) genetic correlations between US_BF and BF_AREA. Table 3 Estimates of heritabilities (h 2 ), common litter effects (c 2 ), along with additive (σ 2 a), common litter (σ 2 c) and phenotypic variances (σ 2 p), from univariate analyses Trait N Mean h 2 c 2 σ 2 a σ 2 c σ 2 p IGF-I ± ± FI_TOT ± ADG ± ± ADG ± ± FCR ± US_BF ± ± M_AREA ± BF_AREA ± PH1_K ± IMF ± BELLY ± Table 4 Phenotypic variances for IGF-I (σ 2 p1) and the performance trait (σ 2 p2), along with genetic (ra), common litter (rc), residual (re) and phenotypic correlations (rp) between IGF-I and performance traits, from bivariate analyses Trait σ 2 p1 σ 2 p2 ra rc re rp FI_TOT ± ± ±0.05 ADG ± ± ± ±0.05 ADG ± ± ± ±0.05 FCR ± ± ±0.05 US_BF ± ± ± ±0.05 M_AREA ± ± ±0.05 BF_AREA ± ± ±0.05 PH1_K ± ± ±0.05 IMF ± ± ±0.06 BELLY ± ± ±0.05 AGBU Pig Genetics Workshop March

6 Genetic correlations between IGF-I and BELLY or M_AREA are high and negative. Genetic correlations between IGF-I and the remaining traits are very low (PH1_K and ADG3) to moderate (ADG2 and IMF). The relatively low genetic correlations between IGF-I and growth are consistent with the average of estimates from previous trials. As with some other trials where performance testing was to a high average end weight genetic correlations between growth while on test and juvenile IGF-I were in a favourable direction. Residual correlations between IGF-I and the performance traits tended to be low, and not significantly different from zero. This result is consistent with traits measured at very different times (~ 5 months apart) during an animal s life. Overall, phenotypic correlations were positive, and significantly different from zero, between IGF-I and FI_TOT, FCR, US_BF and BF_AREA, while they were negative for M_AREA and BELLY. 3. Conclusions Genetic and phenotypic correlations indicate that animals with lower juvenile IGF-I should have lower feed intake, feed conversion ratio, backfat and back fat area, and higher muscle area and belly score. Results from this trial are overall consistent with those from previous trials, supporting a downward selection strategy for juvenile IGF-I. However, relatively few records available for analyses contributed to the large standard errors for all parameter estimates. Subsequently, it is recommended for implementation to use estimates pooled across trials, where possible. Acknowledgments Presentation of results for the trial conducted by Mitteldeutscher Schweinezuchtverband e. V. (MSZV) was made possible by MSZV and Primegro Limited, owners of the exclusive license to the patented PrimeGRO IGF-1 technology. Further Reading Bunter, K., Hermesch, S., Luxford, B.G., Lahti, K. and Sutcliffe, E. (2002). IGF-1 concentration measured in juvenile pigs provides information for breeding programs: a mini review. Proceedings of the 7 th World Congress on Genetics Applied to Livestock Production, Session 03: Pig breeding, Communication No AGBU Pig Genetics Workshop March